What do agents learn?
	• If-then decision structures (decision trees) • Function approximation (neural networks) • Action (control) policy (reinforcement learning) • etc. à lots of things! • Learning agents have at least one adaptive component of their architecture.